Should CIOs Use AI For Hiring?

CIOs need to make AI less mysterious when it comes to hiring
CIOs need to make AI less mysterious when it comes to hiring
Image Credit: Steve Jurvetson

The hiring process at any company is a complex undertaking. CIOs are often involved in the mechanics of how the process works. With the arrival of artificial intelligence (AI) technology, AI is now being incorporated into many of the tools that companies are using as a part of their hiring process. What this means is that from the outside, how the hiring process actually works has started to become somewhat of a mystery. It is now the CIOs job to make sure that we understand how the process works and to remove the mystery that is surrounding it.

Disclosing The Use Of AI In The Hiring Process

CIOs know that artificial-intelligence tools are seeing ever broader use in hiring. However, this practice is also being hotly criticized because we rarely understand how these tools select candidates, and whether the candidates they select are, in fact, better qualified than those who were rejected. To help answer these crucial questions, CIOs should give job seekers more information about the hiring process and the decisions being made. A solution that is being proposed is a twist on something we see every day: nutritional labels. Specifically, CIOs would provide job candidates with simple, standardized labels that show the factors that go into the AI’s decision.

How would a CIO implement this? When people apply for a job, they will now see a list of the hiring criteria, such as degree requirements, specific skills and the number of years of experience, so that they know precisely what the company is looking for. This way, if the applicant is rejected, the AI will present them with another list, showing where they didn’t meet the criteria or compared unfavorably to other applicants – basically the reasoning behind the decision. In other words, CIO’s should show people very clearly what factors are used to judge them, just as we show people the ingredients that go into their food. Let’s face it – we desperately need such a system. AI’s new widespread use in hiring far outpaces our collective ability to keep it in check. We are struggling to understand, verify and oversee it. Is a résumé screener able to identify promising candidates, or is it picking up irrelevant, or even discriminatory, patterns from historical data? Is a job seeker participating in what any of us would call a fair competition if he or she is unable to pass an online personality test, despite having other qualifications needed for the job?

The “posting label” would be a short, simple and clear set of requirements that an AI screener will be looking for in applicants. An example of this would be a posting label for an art-director position that might list “B.S. in communications or similar,” “two years of full-time experience” and “expert knowledge of Adobe Design Suite.” The posting label at the same time would explain the assessment process. Will the AI consider only submitted résumés, or will it also use applicants’ public LinkedIn profiles and Twitter feeds? Will it use their credit histories? Will any video interview be required? Which parts of the candidate’s application are processed by a machine and which are processed by a human?

Explaining The Choice That AI Makes

CIOs understand that having clear criteria for decisions not only helps applicants but it can also give employers vital information. Many times, an AI will make judgment calls that are opaque. CIOs often don’t know what data AI screeners are using, or how they analyze that data to make a final decision. The new labels can show managers the factors that the AI is using to screen applicants – and let those managers decide if those factors need to be changed in order to get better candidates. For instance, does the AI need to be given more or perhaps different training data, covering different job roles and demographic groups, in order to avoid making biased and arbitrary decisions? Likewise, what is the predictive accuracy of the tool for different demographic groups that the company is interested in hiring from? What features of past applicants’ profiles led to a positive or a negative decision by the tool, and can job relevance of these features be substantiated in any way?

One concern that CIOs may have is if these labels will motivate strategic manipulation or “gaming.” However, today there is already strategic manipulation happening in the hiring process: career services routinely offer training and advice on how to make a résumé attractive to the algorithmic tools that will be processing résumés. It is hoped that more transparency will help alleviate unproductive gaming and tilt the balance in favor of positive change, motivating candidates to actually improve their qualifications, rather than to make it seem like they are qualified. Humans – and not an AI – should ultimately be the ones making the final call on whom to hire. But, like it or not, many CIOs use AI systems at different stages of the hiring process – and that practice is only going to become more common. If CIOs and their companies are relying on AI, those tools should be as transparent as possible, and job seekers should be permitted to have a say in how their data is used.

Ultimately, I think that we can all agree that hiring is a complex process. This is a multistep process in which we trade off objective criteria, such as an applicant’s degree requirements and measurable skills, against subjective factors such as how well we think that they will fit into the team and pick up new skills. We can bring in AI to help us alleviate some of this complexity. However, we cannot forget that AI tools work to our specifications, and they do best when those specifications are clear. It is possible to use AI effectively for parts of the hiring process. A good example is to identify clear requirements-based matches. But AI tools cannot exercise human discretion or apply human subjective judgment. The hope is that the use of labels will help CIOs come to a consensus on which decisions they should leave to an AI, and which they should make themselves.

What All Of This Means For You

One of the most important jobs that is done at any company is the hiring of new workers. The process of identifying, interviewing, and then onboarding new talent is something that can be complex. More and more firms are turning to new tools that incorporate Artificial Intelligence (AI) technology in order to do a better job of making good decisions about potential job applicants. However, the way that the AI tools evaluate candidates can be mysterious. Is there anything that CIOs can do to help make things clearer for everyone who is involved?

AI tools can take inputs and produce outputs and its possible that nobody will understand how this is being done. CIOs should give job seekers more information about the hiring process and the decisions. CIOs can implement a system which will show the candidate a list of the hiring criteria. If the applicant is rejected, the AI will present them with another list, showing where they didn’t meet the criteria. CIOs understand that having clear criteria for decisions not only helps applicants but it can also give employers vital information. CIOs need to understand how AI tools are going about making their decisions. CIOs need to make sure that the specifications being used by their AI tools are clear.

Hiring the right people to work at their company is a task that the CIO has to provide assistance with. The use of technology in this task can allow it to move quicker and hopefully allows the company to hire better candidates. AI technology is well suited to helping to streamline the hiring process. However, CIOs have a responsibility to make sure that everyone involved in the hiring process can understand how the AI tools are making their decisions. If they can do this, then CIOs will have helped their company to do a better job of hiring the right people.

– Dr. Jim Anderson Blue Elephant Consulting –
Your Source For Real World IT Department Leadership Skills™

Question For You: Do you think that CIOs should tell candidates that they will be evaluated by AI tools?

Click here to get automatic updates when The Accidental Successful CIO Blog is updated.
P.S.: Free subscriptions to The Accidental Successful CIO Newsletter are now available. Learn what you need to know to do the job. Subscribe now: Click Here!

What We’ll Be Talking About Next Time

On top of everything else that a CIO is being asked to do, we are also in charge of getting the company’s employees to do their part in order to secure the company’s IT assets. It turns out that this is actually trickier than it may initially seem to do. You have to understand who you are dealing with and you need to have a clear vision for what you want them to be doing. Once you know all of this, you’ll need a plan for getting your employees to do the right thing. Oh, and trying to scare them into doing the right thing is probably not the way to go about doing it.